This is a basic research proposal concerning the process by which a cognitive factor--attributional style--produces emotional and achievement deficits in normal children following bad events. This is not a clinical study of disturbed persons, even though its findings would be important for the etiology of affective disorders. This project has four major goals: 1) to predict which children will show achievement problems in school work following bad events; 2) to predict which children will suffer depressive affect and behavior following bad events; 3) to investigate the causal relationships among attributional style, bad life events and social support on achievement and depressive deficits; and 4) to investigate how children acquire an insidious attributional style which in turn can undermine mental health. Toward these ends 400 children in the third grade of the Princeton, New Jersey, area elementary schools and their parents will participate in a four-year longitudinal study. At the outset of the study, the children and their parents will receive baseline assessments for attributional style, social support, recent life events, and various demographic variables. In addition, school achievement test scores, grades, teacher ratings and depressive affect will be assessed. Then, at twice-a-year intervals for each of the next four years the children will be reassessed on these same variables. Cross-sectional analyses from each assesement will be made, and longitudinal analyses using modern techniques of causal modeling will be performed on the predictive data. At the end of the study, we hope to be able to identify in advance which children will be most vulnerable to achievement and emotional problems when they encounter loss.